22 October 2019 in Human Brain Mapping
Resting‐state fMRI has shown the ability to predict task activation on an individual basis by using a general linear model (GLM) to map resting‐state network features to activation z‐scores. The question remains whether the relatively simplistic GLM is the best approach to accomplish this prediction. In this study, several regression‐based machine‐learning approaches were compared, including GLMs, feed‐forward neural networks, and random forest bootstrap aggregation (bagging). Resting‐state and task data from 350 Human Connectome Project subjects were analyzed. First, the effect of the number of training subjects on the prediction accuracy was evaluated. In addition, the prediction accuracy and Dice coefficient were compared across models. Prediction accuracy increased with the training number up to 200 subjects; however, an elbow in the prediction curve occurred around 30–40 training subjects. All models performed well with correlation matrices, which displayed correlation between actual and predicted task activation for all subjects, exhibiting a strong diagonal trend for all tasks. Overall, the neural network and random forest bagging techniques outperformed the GLM. These approaches, however, require additional computing power and processing time. These results show that, while the GLM performs well, resting‐state fMRI prediction of task activation could benefit from more complex machine learning approaches.
25 March 2019 in Scientific Reports
Breath holding (BH) is a viable vasodilatory stimulus for calculating functional MRI-derived cerebral vascular reactivity (CVR). The BH technique suffers from reduced repeatability compared with gas inhalation techniques; however, extra equipment is needed to perform gas inhalation techniques, and this equipment is not available at all institutions. This study aimed to determine the sensitivity and repeatability of BH activation and CVR using a multiband multi-echo simultaneous arterial spin labelling/blood oxygenation level dependent (ASL/BOLD) sequence. Whole-brain images were acquired in 14 volunteers. Ten subjects returned for repeat imaging. Each subject performed four cycles of 16 s BH on expiration interleaved with paced breathing. Following standard preprocessing, the echoes were combined using a T2*-weighted approach. BOLD and ASL BH activation was computed, and CVR was then determined as the percent signal change related to the activation. The “M” parameter from the Davis Model was also computed by incorporating the ASL signal. Our results showed higher BH activation strength, volume, and repeatability for the combined multi-echo (MEC) data compared with the single-echo data. MEC CVR also had higher repeatability, sensitivity, specificity, and reliability compared with the single-echo BOLD data. These data support the usefulness of an MBME ASL/BOLD acquisition for BH CVR and M measurements.
1 January 2021 in NeuroImage
Recent advances in functional MRI techniques include multiband (MB) imaging and multi-echo (ME) imaging. In MB imaging multiple slices are acquired simultaneously leading to significant increases in temporal and spatial resolution. Multi-echo imaging enables multiple echoes to be acquired in one shot, where the ME images can be used to denoise the BOLD time series and increase BOLD sensitivity. In this study, resting state fMRI (rs-fMRI) data were collected using a combined MBME sequence and compared to an MB single echo sequence. In total, 29 subjects were imaged, and 18 of them returned within two weeks for repeat imaging. Participants underwent one MBME scan with three echoes and one MB scan with one echo. Both datasets were processed using standard denoising and advanced denoising. Advanced denoising included multi-echo independent component analysis (ME-ICA) for the MBME data and ICA-AROMA for the MB data. Resting state functional connectivity (RSFC) was evaluated using both selective seed-based and whole grey matter (GM) region-of-interest (ROI) based approaches. The reproducibility of connectivity metrics was also analyzed in the repeat subjects. In addition, functional connectivity density (FCD), a data-driven approach that counts the number of significant connections, both within a local cluster and globally, with each voxel was analyzed. Regardless of the standard or advanced denoising technique, all seed-based RSFC was significantly higher for MBME compared to MB. Much more GM ROI combinations showed significantly higher RSFC for MBME vs. MB. Reproducibility, evaluated using the dice coefficient was significantly higher for MBME relative to MB data. Finally, FCD was also higher for MBME vs. MB data. This study showed higher RSFC for MBME vs. MB data using selected seed-based, whole GM ROI-based, and data-driven approaches. Reproducibility found also higher for MBME data. Taken together, these results indicate that MBME is a promising technique for rs-fMRI.
1 February 2018 in PLOS ONE
Typical simultaneous blood oxygenation-level dependent (BOLD) and arterial spin labeling (ASL) sequences acquire two echoes, one perfusion-sensitive and one BOLD-sensitive. However, for ASL, spatial resolution and brain coverage are limited due to the T1 decay of the labeled blood. This study applies a sequence combining a multiband acquisition with four echoes for simultaneous BOLD and pseudo-continuous ASL (pCASL) echo planar imaging (MBME ASL/BOLD) for block-design task-fMRI. A multiband acceleration of four was employed to increase brain coverage and reduce slice-timing effects on the ASL signal. Multi-echo independent component analysis (MEICA) was implemented to automatically denoise the BOLD signal by regressing non-BOLD components. This technique led to increased temporal signal-to-noise ratio (tSNR) and BOLD sensitivity. The MEICA technique was also modified to denoise the ASL signal by regressing artifact and BOLD signals from the first echo time-series. The MBME ASL/BOLD sequence was applied to a finger-tapping task functional MRI (fMRI) experiment. Signal characteristics and activation were evaluated using single echo BOLD, combined ME BOLD, combined ME BOLD after MEICA denoising, perfusion-weighted (PW), and perfusion-weighted after MEICA denoising time-series. The PW data was extracted using both surround subtraction and high-pass filtering followed by demodulation. In addition, the CBF/BOLD response ratio and CBF/BOLD coupling were analyzed. Results showed that the MEICA denoising procedure significantly improved the BOLD signal, leading to increased BOLD sensitivity, tSNR, and activation statistics compared to conventional single echo BOLD data. At the same time, the denoised PW data showed increased tSNR and activation statistics compared to the non-denoised PW data. CBF/BOLD coupling was also increased using the denoised ASL and BOLD data. Our preliminary data suggest that the MBME ASL/BOLD sequence can be employed to collect whole-brain task-fMRI with improved data quality for both BOLD and PW time series, thus improving the results of block-design task fMRI.
30 November 2019 in Journal of Magnetic Resonance Imaging